Asuswrt-Merlin 3.0.0.4.374.33 Beta 6 available

[RMerlin]

Howdy folks,

Quick beta refresh, Beta 6 is now available.

What I need tested in this build:

• AiCloud (the crashing issues of Beta 5 should be fixed)
• minidlna scanning (I re-enabled Asus's thumbnail scanner)

If you get any error while scanning with minidlna please post the log output. Some errors are perfectly normal, and won't prevent minidlna to continue scanning the rest of your database (for instance, when trying to process an image that requires more RAM than the router has). The key is to ensure that the scanning process does complete normally.

One way to see what minidlna is doing is to install strace through Entware or Optware, then use it to monitor the activity of the main minidlna process. Determine what its PID is (through "ps w"), then:

strace -p pid_value

Most of it might be gibberish, but in there you will also see the filename of the file is just processed.

You can cancel that monitoring using Ctrl-C.

I know that there is newer GPL code out, but the plan for 374.33 is just to finalize testing of what's already in it, and keep the newer GPL stuff for the next release. There's some new features in that latest GPL (Configurable WAN watchdog amongst other things) that will probably require testing of its own.


Thanks everyone!

 

[connorm]

just installed this, i will let you know if there is any issues i run into, but i really wish the problem with upnp overiding default port forwarding rules would be fixed because i have to leave upnp off for now and eventually if this isnt fixed i will have to go to other firmware, and i have verified this issue does not happen on my d link router so it could be an a small asus problem (only the rt-ac66u or just my unit) a big asus problem(every asus router has this issue) a firmware issue (could be just merlinwrt cause never happened to me on asuswrt)

 

[krum]

Merlin
I will stick with beta 5 and continue to test parental controls
Will keep you posted

Chris

 

[RMerlin]

just installed this, i will let you know if there is any issues i run into, but i really wish the problem with upnp overiding default port forwarding rules would be fixed because i have to leave upnp off for now and eventually if this isnt fixed i will have to go to other firmware, and i have verified this issue does not happen on my d link router so it could be an a small asus problem (only the rt-ac66u or just my unit) a big asus problem(every asus router has this issue) a firmware issue (could be just merlinwrt cause never happened to me on asuswrt)

UPNP rules are currently processed before manual forwards. I will have to see if it can easily be moved at the end without creating other issues.

 

[NightOwl326]

Why is .33 beta 5 SDK5 the only Merlin firmware that when accessing Netflix on my TV I get an "no internet available message" (only with
Netflix and internet is connected for all other applications such as YouTube) and rolling back to back to beta 1 and all is good. Reflashed a couple of times and cleared settings and entered manually and no Netflix. ???

 

[highwire]

Why is .33 beta 5 SDK5 the only Merlin firmware that when accessing Netflix on my TV I get an "no internet available message" (only with
Netflix and internet is connected for all other applications such as YouTube) and rolling back to back to beta 1 and all is good. Reflashed a couple of times and cleared settings and entered manually and no Netflix. ???

Co-incidence. There were issues with Netflix this evening. Sometimes it would time-out, other times it would connect, sometimes with SuperHD support, and sometimes without.

 

[NightOwl326]

Co-incidence. There were issues with Netflix this evening. Sometimes it would time-out, other times it would connect, sometimes with SuperHD support, and sometimes without.

Been fussing with this for 3 days, its the firmware. Back and forth multiple times, Beta 1 connects, Beta 5 no connects....

 

[PrivateJoker]

Since the wide roll out of "SuperHD" last week Netflix has brought Akami into their existing infrastructure of AWS & proprietary server CDN. If I look at my connection logs, certain devices (Apple TV) will actually draw from all 3 of those resources at the same time, others (Tivos, only one or two). They are probably ironing some kinks out.

 

[connorm]

UPNP rules are currently processed before manual forwards. I will have to see if it can easily be moved at the end without creating other issues.

thank you, right now i have UPNP off and both xbox manually forwarded to the ports that upnp decides to forward to them (upnp gives 1 xbox 3074, and another 19059 and both open nat) 3074 to my xbox and 19059 to my brothers, mine is open nat, but his is not, why the same ports being forwarded via upnp will give open nat i do not understand

 

[RMerlin]

Some Netflix users also reported having trouble with Netflix if they had IPv6 enabled (via a tunnel or otherwise).

 

[damir]

yea, that was reported here as well:
http://www.linksysinfo.org/index.php?threads/warning-ipv6-and-netflix.69097/

 

[GoNz0]

Pleased to report that DLNA is now scanning and I can connect to my USB devices

 

[Quiller]

Asuswrt-Merlin 3.0.0.4.374.33 Beta 6

My problems with the AiCloud app for iOS have been resolved. great job.

 

[JGrana]

Minidlna still out of memory Photos

Sorry to report that Beta 6, minidlna is still reporting malloc failures and out of memory on a large number of Photos...

First a stream of these messages:

[2013/10/02 06:41:03] image_utils.c:419: warn: malloc failed

Then these:
[2013/10/02 06:42:15] sql.c:41: error: SQL ERROR 7 [out of memory]

free reports plenty of buffers and swap hasn't been exhausted...

I will try the /opt/bin version you posted last week. I suspect this will work fine.

Good luck!!!

 

[SN_]

Here an interesting observation with 33_beta6-sdk5 that may help to track down the low txpower problems:
With this beta i do have the same drop in txpower for 2.4GHz as with sdk6 based FW. Setting txpower is not possible at all and everything related to txpower seems to have the same dysfunction as with a SDK6 based FW!

'wl ver' returns:
5.100 RC138.20
wl0: Jan 23 2013 15:25:28 version 5.100.138.20

With 33_beta5b-sdk5 txpower was just fine.

Can someone confirm this?

 

[SN_]

Cant reproduce the issue. After downgrading, making a hard reset, uploading configuration and upgrading to beta6-sdk5 again everything runs fine now.

 

[JGrana]

Sorry to report that Beta 6, minidlna is still reporting malloc failures and out of memory on a large number of Photos...

First a stream of these messages:

[2013/10/02 06:41:03] image_utils.c:419: warn: malloc failed

Then these:
[2013/10/02 06:42:15] sql.c:41: error: SQL ERROR 7 [out of memory]

free reports plenty of buffers and swap hasn't been exhausted...

I will try the /opt/bin version you posted last week. I suspect this will work fine.

Good luck!!!

I can confirm that the minidlna version Merlin posted last week did complete all scans fine. For info, there are 55,629 photos.

 

[RMerlin]

Sorry to report that Beta 6, minidlna is still reporting malloc failures and out of memory on a large number of Photos...

First a stream of these messages:

[2013/10/02 06:41:03] image_utils.c:419: warn: malloc failed

Then these:
[2013/10/02 06:42:15] sql.c:41: error: SQL ERROR 7 [out of memory]

free reports plenty of buffers and swap hasn't been exhausted...

I will try the /opt/bin version you posted last week. I suspect this will work fine.

Good luck!!!

This error isn't a bug, simply that a specific photo minidlna tried to process would require more memory than available. The code shows that minidlna will report the failure, and simply proceed with the next file without any problem:

image_s *
image_new(int32_t width, int32_t height)
{
        image_s *vimage;

        if((vimage = (image_s *)malloc(sizeof(image_s))) == NULL)
        {
                DPRINTF(E_WARN, L_METADATA, "malloc failed\n");
                return NULL;
        }
        vimage->width = width; vimage->height = height;

        if((vimage->buf = (pix *)malloc(width * height * sizeof(pix))) == NULL)
        {
                DPRINTF(E_WARN, L_METADATA, "malloc failed\n");
                free(vimage);
                return NULL;
        }
        return(vimage);
}

So, this isn't a bug in need of fixing, it's just a harmless limitation. I initially thought the scanner was stopping scanning altogether, but that isn't the case.

 

[JGrana]

Minidlna Testing

So I really should be testing that the beta 6 minidlna, even though it throws memory faults continues.

Ok, easy enough. From what I recalled last time, it bailed out on the 2007 folder (I have them organized by year). I will go back to the built in version and let it finish - then check with a client later today.

I will let you know this evening.

Thanks again.

 

[Mr Zippy]

Apologies for jumping in late, only just found this forum as I'd previous emailed direct with issues to Merlin...

Anyway I'm asking a silly question I know but since the SDK6 firmware for the RT-N66U I have 2 strange 2.4GHz issues:

1. If I manually select a channel I can't set to 20/40 bandwidth as doing this resets the control channel to auto.
2. Why are the channels now suffixed with an 'l' or a 'u'?

Thanks

Neil